AI+ New Calling: A New Start for Voice Service

To develop innovative services for New Calling, ZTE has continuously enhanced the IMS architecture of the foundational real-time communication network by incorporating data channels, capability exposure, applets, and AI technologies.

In terms of AI technology support, ZTE and China Mobile jointly launched media-plane plug-in technology to meet the low latency, high reliability, and easy scalability requirements for intelligent multimodal conversion and processing of innovative services. This enables the online deployment of AI capability through an agile base solution, facilitating the rapid rollout of innovative services.

As shown in Fig. 1, the media-plane plug-in architecture is divided into four layers: service, framework, capability plug-in, and support. The service, framework, and support layers provide the system’s basic functions, while the capability layer is an optional function that can be loaded as required.

• Service layer: Handles media service interfaces and converts them into plug-in orchestration.

• Framework layer: Contains the pipeline engine, loads plug-ins according to the orchestration, and executes the pipeline.

• Capability plug-in layer: Supports both self-owned and third-party plug-ins, include basic media processing plug-ins (e.g., codec), multimode media processing plug-ins (e.g., XR), and AI processing plug-ins (e.g., face and intent recognition).

• Support layer: Provides the runtime environment for plug-ins.

The architecture decouples plug-in capabilities from the system. With system stability as a prerequisite, new capabilities and functions can be rapidly deployed and launched through flexible plug-in orchestration. Additionally, based on  standardized plug-in architecture and interfaces, ZTE promotes collective innovation and ecosystem co-construction of AI and new media processing capabilities, facilitating the development of AI-powered New Calling.

The innovative agile base solution integrates calling and AI to deliver a brand-new calling experience, expand the scope of communication services, and create a new commercial closed loop.

AI-generated content (AIGC) greatly reduces the threshold and cost of content creation, offering rich and personalized displays for individual and enterprise users through the New Calling feature “lighting up screen”.

The AI agent-based call assistant identifies user intent during a call and generates relevant images and videos to enhance the interaction. For example, when a user talks about a tourist attraction or a product feature, the assistant can display related visuals on the terminal screen in real time. The assistant can also provide intelligent responses to avoid missing calls. For calls from relatives, friends, and colleagues, it can create personalized effects for emotional engagement. For unknown callers, it helps prevent harassment and fraud.

Moreover, the call assistant can offer services such as food ordering, travel booking, and financial assistance via third-party applications, enabling closed-loop value-added services during calls.

For industrial applications, such as service marketing, customer care, and after-sales guarantee, AI+ New Calling offers greater efficiency and lower costs. For example, AI intelligent customer service provides users with faster, more personalized experiences, while remote guidance such as screen sharing and AR marker greatly improve service efficiency and reduce service costs.

Practice and Prospects

Leveraging advancements in 5G-A and AI technologies, China Mobile, together with ZTE and other vendors, has taken the lead in researching, testing, verifying, and promoting New Calling services, effectively enhancing call engagement and service quality, and driving steady user growth and sound business development.

While mainstream Chinese operators are actively conducting pilot projects and commercial deployments of New Calling services, international operators are also showing strong interest and have begun pilot verifications. The global expansion of New Calling services is underway.

Demand for new services and technological advancement are driving the continuous evolution of real-time communication networks. Leading Chinese operators have carried out research on the next-generation real-time communication network, and initiated related requirements projects in ITU-T. ZTE is actively participating in discussions on requirements and architecture for operators' next-generation real-time communication networks, contributing to the preparation and release of white papers.

The next-generation real-time communication network architecture will feature native AI capabilities, enabling intelligence across content production, network control, and media processing, supporting immersive and virtual-real integrated service experiences, and ultimately helping operators speed up commercial monetization.